Searching for relevant software change artifacts using semantic networks

The discovery of software artifacts (files, documents, and datasets) relevant to a change request, can increase software reuse and reduce the cost of software development and maintenance. However, traditional search techniques often fail to provide the relevant documents because they do not consider relationships between software artifacts. We propose the creation of Semantic Networks which convey such relationships and assist in automatically discovering not only the requested artifacts based on a user query, but additional relevant ones that the user may not be aware of. Subsequently, we increase the accuracy of the returned artifacts by applying appropriate contexts. Experimental results show that this approach leads to better recall and precision compared to existing full-text search approaches.